Breast cancer is the most common malignancy in women. Mutations in <i>BRCA1</i> and <i>BRCA2</i> genes play an important role in the development of early-onset familial breast cancer. Our goal is to carry out functional dissection of these genes using a mouse model system. Since complete loss of function mutations result in embryonic lethality, we are engaged in generating an array of specific mutations along the length of proteins encoded by these genes. To accomplish this, we have developed a transgenic mouse model system that combines the use of existing <i>Brca1</i> and <i>Brca2</i>knockout mice and the use of bacterial artificial chromosomes (BAC) containing these genes. Desired mutations can be generated in the <i>Brca1</I> or <i>Brca2</i> gene in the BAC and the phenotypic effect of the mutation can be analyzed in transgenic mice that are homozygous mutant for that gene. In a continuing effort to understand the role of BRCA1 and BRCA2 in various biological processes, we have developed humanized mouse models for functional analysis of human <i>BRCA1</i> and <i>BRCA2</i> genes. We have shown that wild-type human <i>BRCA1</i> and <i>BRCA2</i> genes under the control of their own promoter present in BAC clones are fully functional in mice and can rescue the lethality associated with loss of function mutations of the respective genes in mice. The humanized mouse models provide an experimentally tractable system to generate mutations identified in human breast cancer patients and analyze how they result in tumorigenesis. We have developed a simple but powerful method to generate subtle alterations in the BACs by using recombineering in bacteria. This method is based on the generation of recombinants in the BAC DNA by using oligonucleotides as targeting vectors with as few as 70 bases of homology. We have demonstrated that this technique can be used to generate single base changes, small insertions and deletions. Due to the high efficiency of recombineering, recombinant clones can be identified by a simple PCR based screening method without the use of any selectable marker. We have used this method to generate several cancer causing missense mutations in <i>BRCA1</i> and <i>BRCA2</i> in the BAC and are examining their phenotypic consequences in mice in a <i>Brca1</i> or <i>Brca2</i> mutant background. Based on the expression of BRCA2 on meiotic chromosomes, it has been predicted to be required for normal meiotic progression. Since <i>Brca2</i> null mutants are embryonic lethal, its precise function in meiosis is unknown. We have generated a mouse model to describe a role for BRCA2 in meiosis. These mice lack the endogenous <i>Brca2</I> gene but lethality is rescued by a human <i>BRCA2</i> transgene. However, these mice are sterile due to poor expression of the transgene in testes and ovaries. We have used these mice to describe a role for BRCA2 in meiosis that is sexually dimorphic. We have found that the BRCA2-deficient spermatocytes undergo apoptosis unlike the BRCA2-deficient breast and ovarian epithelial cells that undergo neoplastic growth. This suggests that the cellular response to DNA damage may be tissue specific, which may explain why mutations in this widely expressed gene do not result in cancer in all tissues. A mutation in an evolutionarily conserved domain of BRCA2 has revealed a novel function for the protein in alkyl-DNA repair. Mutant mice develop normally but embryonic fibroblasts derived from these mice are extremely sensitive to N-methyl-N'-nitro-N-nitroso-guanidine (MNNG), an alkylating agent, which results in the conversion of guanines to O6-methylguanines. The major pathway of repair of this methylated guanine involves removal of a methyl group by the O6-methylguanine-methyl transferase (MGMT) enzyme. Our current research is focused on understanding the interaction between BRCA2 and MGMT function as it may open new avenues for breast cancer therapy. Although transgenic humanized mice using BACs are excellent <i>in vivo</i> models for understanding the functional significance of the human mutations, they are not ideally suited for characterizing large number of mutations. Hence, we are developing an embryonic stem (ES) cell-based functional assay to rapidly screen large number of mutations. The assay is based on the observation that mouse ES cells that are mutated in both alleles of either <i>Brca1</i> or <i>Brca2</i> genes, do not survive. We intend to use this assay system to study large number of mutations in a short time by testing their ability to rescue the ES cell lethality. Those mutations that can rescue the lethality will be examined for defects in various DNA repair processes. We have also initiated the functional analysis of <i>Rad51l2</i>, a member of the <i>Rad51</i>-like family. Rad51 is one of the key DNA repair protein that interacts with both BRCA1 and BRCA2. We plan to determine whether RAD51L2 interacts with BRCA1 or BRCA2 proteins as <i>in vitro</i> studies have demonstrated a role for RAD51L2 in homologous DNA repair